{
  "slug": "plover-inspectable-plans-for-gui-agent-repair-715193",
  "title": "Plover: Inspectable Plans for GUI Agent Repair",
  "dek": "Plover makes GUI-agent plans persistent, visible, and editable so users can inspect execution, guide replanning, and repair failures without discarding prior progress.",
  "summary": "Plover exposes persistent editable GUI-agent plans so users can inspect, guide, and repair execution in dynamic interfaces.",
  "tags": [
    "GUI automation",
    "vision-based agents",
    "multimodal agents",
    "plan-centric interaction",
    "replanning",
    "agent supervision",
    "agent repair",
    "human-computer interaction"
  ],
  "published_at": "2026-07-19T01:08:32.962+00:00",
  "grade": 8.4,
  "agent_utility": 8,
  "price_usdc": 0.271,
  "stats": {
    "claims": 16,
    "entities": 25,
    "faq": 8,
    "downloads": 0,
    "paid_downloads": 0
  },
  "preview": {
    "faq_questions": [
      "What is Plover?",
      "What problem does Plover address?",
      "How is Plover different from typical vision-based GUI agents?",
      "What kinds of intervention does Plover support?",
      "Does Plover restart tasks after failure?",
      "How was Plover evaluated?",
      "What quantitative results are available from the abstract?",
      "Does the abstract identify the models or benchmarks used?"
    ],
    "entity_names": [
      {
        "name": "Plover",
        "type": "system"
      },
      {
        "name": "Graphical user interface automation",
        "type": "domain"
      },
      {
        "name": "Real-world GUI environments",
        "type": "environment"
      },
      {
        "name": "Dynamic layouts",
        "type": "failure factor"
      },
      {
        "name": "Unexpected dialogs",
        "type": "failure factor"
      },
      {
        "name": "Evolving interface states",
        "type": "failure factor"
      },
      {
        "name": "Autonomous agents",
        "type": "agent class"
      },
      {
        "name": "Vision-based multimodal agents",
        "type": "agent class"
      },
      {
        "name": "Screenshots",
        "type": "input modality"
      },
      {
        "name": "Natural-language instructions",
        "type": "input modality"
      },
      {
        "name": "Task plans",
        "type": "artifact"
      },
      {
        "name": "Replanning",
        "type": "process"
      },
      {
        "name": "Planner-executor architecture",
        "type": "architecture"
      },
      {
        "name": "Explicit supervision",
        "type": "interaction mechanism"
      },
      {
        "name": "Localized correction",
        "type": "interaction mechanism"
      },
      {
        "name": "Editable plans",
        "type": "artifact"
      },
      {
        "name": "Natural-language guidance",
        "type": "interaction mechanism"
      },
      {
        "name": "Screenshot-grounded interventions",
        "type": "interaction mechanism"
      },
      {
        "name": "Prior progress",
        "type": "execution state"
      },
      {
        "name": "Formative study participants",
        "type": "study population"
      },
      {
        "name": "Benchmark failure-case repair",
        "type": "evaluation method"
      },
      {
        "name": "Scenario-based workflow analyses",
        "type": "evaluation method"
      },
      {
        "name": "Transparent GUI automation",
        "type": "quality attribute"
      },
      {
        "name": "Controllable GUI automation",
        "type": "quality attribute"
      },
      {
        "name": "Adaptable GUI automation",
        "type": "quality attribute"
      }
    ],
    "related_work_titles": [
      "Recent vision-based multimodal agents",
      "GUI agents with internal planning and adaptation",
      "Supervised GUI automation approaches",
      "Benchmark failure-case repair for GUI agents"
    ],
    "application_industries": [
      "GUI automation",
      "AI agent infrastructure",
      "Robotic process automation",
      "Human-computer interaction",
      "Workflow automation"
    ],
    "glossary_terms": [
      "Graphical user interface",
      "GUI automation",
      "Vision-based multimodal agent",
      "Task plan",
      "Persistent artifact",
      "Replanning",
      "Planner-executor architecture",
      "Explicit supervision",
      "Localized correction",
      "Editable plan",
      "Natural-language guidance",
      "Screenshot-grounded intervention",
      "Drift from user intent",
      "Structurally repairable",
      "Benchmark failure-case repair",
      "Scenario-based workflow analysis"
    ]
  },
  "purchase": {
    "protocol": "x402",
    "version": 2,
    "network": "eip155:8453",
    "asset": "USDC",
    "price": 0.271,
    "full_markdown_url": "https://x402.disruptive-concepts.com/api/public/article/plover-inspectable-plans-for-gui-agent-repair-715193/full",
    "full_json_url": "https://x402.disruptive-concepts.com/api/public/article/plover-inspectable-plans-for-gui-agent-repair-715193/full?format=json"
  }
}